We have used human neuroblastoma as a model system to study the molecular events associated with the development of malignant tumors during childhood. Recent work has indicated that among histologically-indistinguishable solid tumors there are different genetic entities which correspond to cells at different stages in the differentiation of the tissue in which solid tumors arise. Such a schema has been hypothesized for both lymphoid and hematopoietic malignancies in which specific markers of differentiation have been extensively characterized, but tumors of solid tissues have previously been classified largely in terms of the organ in which they arise. Our current focus is to use recombinant DNA technology to identify and characterize tissue specific markers, to use these markers to develop clinically efficacious tumor nosologies, and to study the molecular mechanisms important for the regulation of tissue-specific differentiation and the arrest of growth which invariably accompanies it. Ongoing work in this areas suggests that the molecular mechanisms that give rise to different tumor subgroups we have been able to identify are pathologic alterations in regulatory pathways that function during specific times in normal development. During the past year we have expanded our studies in this area and sought profiles of gene expression that distinguish among tumors occurring at several different organ sites. In this manner we hope to subgroup a number of tumor types into biologically homogeneous groups. Ultimately, our goal is to use this information to better classify patients for therapeutic trials and to develop novel therapeutic approaches based on the specific alterations underlying the development of individual malignancies.